References

Libraries for survival analysis

• the core libraries:
  • survival, ggplot for survival curves: survminer, a nice vignette for eha
  • advanced (Machine Learning): Random survival forest, vignette.
    • Installation guide for parallel processing: RSF requires OpenMP (as does data.table), see e.g. OpenMP installation guide.
• the core libraries:
  • lifelines vignettes, scikit-survival
  • Installation notes for scikit-survival on Mac Silicone chips: As per installation guide, we need cmake, so in a terminal brew install cmake, then pip3 install scikit-survival.

Correspondence tables

fct			ex
formula	Surv(durat, fail) ~ x1 + ..
KM estimator	survfit	KaplanMeierFitter
	survminer::ggsurvplot
	arrange_ggsurvplots
Log-Rank Test	survdiff	logrank_test	ex
	pairwise_survdiff
integrated hazard	ggsurvplot(.,fun = “cumhaz”)	naf = NelsonAalenFitter
		naf.plot_cumulative_hazard
		naf.plot_hazard(bandwidth=)
RSF	randomForestSRC::rfsrc	RandomSurvivalForest	ex
PH, Weibull	eha::phreg(., dist = “weibull”)
as AFT	survreg + ConvertWeibull	WeibullAFTFitter	ex
PH, PWE	eha::pchreg(., cuts)	CoxPHFitter( baseline_estimation_method = “piecewise”, )
	survSplit + GLM
PH, partial likelihood	coxph	CoxPHFitter
+ nonpara. baseline haz.	basehaz + locpoly		ex
Penalised Cox regressions	glmnet		ex
PH, propor. test	cox.zph	CoxPHFitter.check_assumptions	ex
	ggcoxzph	proportional_hazard_test
	ggcoxdiagnostics
MPH	parfm

from lifelines import KaplanMeierFitter
from lifelines.statistics import logrank_test
from lifelines import NelsonAalenFitter
from lifelines import CoxPHFitter
from lifelines import WeibullAFTFitter
from lifelines.statistics import proportional_hazard_test

from sksurv.nonparametric import kaplan_meier_estimator
from sksurv.nonparametric import nelson_aalen_estimator
from sksurv.linear_model import CoxPHSurvivalAnalysis

from sksurv.ensemble import RandomSurvivalForest
from sklearn.model_selection import train_test_split
from sklearn.inspection import permutation_importance

The Zen of Python

The principles (import this) are listed as follows :

    Beautiful is better than ugly.
    Explicit is better than implicit.
    Simple is better than complex.
    Complex is better than complicated.
    Flat is better than nested.
    Sparse is better than dense.
    Readability counts.
    Special cases aren't special enough to break the rules.
    Although practicality beats purity.
    Errors should never pass silently.
    Unless explicitly silenced.
    In the face of ambiguity, refuse the temptation to guess.
    There should be one– and preferably only one –obvious way to do it.
    Although that way may not be obvious at first unless you're Dutch.
    Now is better than never.
    Although never is often better than right now.
    If the implementation is hard to explain, it's a bad idea.
    If the implementation is easy to explain, it may be a good idea.
    Namespaces are one honking great idea – let's do more of those!